|
|
|
Wouldn’t it be nice
if we could specify that more than one of the arguments to a method can
participate fully, at runtime, in the dispatch process?
|
|
|
|
Here’s how this
might look in Java. Or might have looked in Java.
|
|
|
|
I gather that with
Java 1.5, the language’s keepers came up with an altogether different role
for angle bracketed type designators.
|
|
|
|
Someone evidently
decided that what Java really needed, to relieve the onerous, evidently
unbearable burden of writing an explicit downcast for Iterator assignments,
was a full blown parametric polymorphic type system, that, when compiled down
to byte code, mysteriously vanishes without a trace.
|
|
|
|
Java now has not one
but two useless type systems as seamlessly integrated as, well, Great Britain
and Northern Ireland / HP and Compaq / Bosnia-Herzegovina…
|
|
|
|
Someone should be
tried for crimes against Computer Science for inflicting this odious
abomination on the body politic. In the hands of the untutored, it promises
to undermine / stand in the way of
everything about Java that had made it possible to engage in the faux-dynamic
style of programming that helped to make many design patterns practical /
popular…
|
|
|